Scientists and clinicians interested in left ventricular microstructure and cardiac contraction.

Scientists and clinicians interested in the field of cardiac diffusion weighted imaging (cDWI) and cardiac diffusion tensor imaging (cDTI).

- The microstructural organistaions of cardiomyocytes and their cyclic re-arrengements throughout the cardiac cycle determine both left ventricular regional contractile function and electrical conduction; and they are subject to remodelling in the presence of disease.

- Improved understanding of altered myocardial microstructural dynamics in patients with cardiac disease may provide new insights into their diagnosis, risk stratification and evaluation of treatment efficacy, as well as potentially identifying new therapeutic targets.

- Cardiac Diffusion Weighted Imaging (cDWI) and cardiac Diffusion Tensor Imaging (cDTI) are non-invasive and non-destructive MRI approaches that can provide information on the three dimensional microstructure of the myocardium.

This presentation will focus on cardiac diffusion weighted imaging (cDWI) and cardiac diffusion tensor imaging (cDTI) for in vivo cardiac tissue characterization.

At the conclusion of this presentation, participants will be better able to:

- Describe the microstructural arrangements of cardiomyocytes in the left ventricular (LV) myocardium.

- Describe the mechanistic basis that underlies thickening of normal LV myocardium in vivo.

- Understand the basic principles of in vivo cDWI and cDTI techniques.

- Understand the potential of in vivo cDWI and cDTI both for basic science research and for clinical research.

The development of contractile dysfunction and adverse left ventricular (LV) remodeling are associated with poor prognosis. The contribution of the micro-architectural arrangement of cardiomyocytes to LV wall thickening, and how it changes through the cardiac cycle, is not widely appreciated as most research is focused at a biochemical, cellular or macroscopic wall thickness scale. Improved understanding of the processes underlying these observations may provide new insights into the risk stratification and personalized treatment planning for these patients, as well as identifying new treatment targets. Furthermore, this may enhance evaluation of novel therapies, aimed at improving clinical outcomes for patients in whom prognosis remains poor despite conventional treatment.

The microstructure of the compact myocardium of the left ventricle (LV) in humans and other mammals consists of a syncytium of cardiomyocytes embedded in a primarily collagen matrix. From LV apex to base, cardiomyocytes progress from left handed helices in the epicardium, through circumferentially arrangements in the mesocardium, into right handed helices in the endocardium [1, 2]. This architecture can be described quantitatively by the helix angle (HA), which transitions transmurally from circa -60o to +60o from epicardium to endocardium [3, 4].

Cardiac contraction involves both longitudinal and circumferential shortening of the LV (~10-25%, depending on direction and depth) together with radial wall thickening (>35%), and twisting of the apex relative to the base [5]. Cardiomyocytes, which are the fundamental contractile element in the heart, only shorten by approximately 15% and thicken by approximately 8% during systole [5].

Cardiomyocytes are organised in laminar microstructures known as sheets or sheetlets [3, 6]. These structures are about 5-10 cardiomyocytes thick, and are separated by collagen-lined shear layers. These sheetlet planes extend both in the direction of local cardiomyocytes long-axis (helical orientation) and across them, in directions that are oblique to the local epicardial wall tangent plane [7]. Sheetlets have been described to swivel cyclically so that they lie more parallel to the local epicardial wall tangent plane in diastole and more perpendicular to it in systole [8], and LV wall thickening and base-to-apex shortening during systole have been attributed mainly to sheetlet reorientation and concomitant shear layer slippage, with just a small contribution from thickening of individual cardiomyocytes.

Diffusion tensor magnetic resonance imaging (DTI) has been most widely applied to in vivo neural microstructure imaging [9-13]. More recently has it also been applied to the study of the myocardium [14-25]. The basic principle is that an MRI signal is attenuated by the self-diffusion of water in the presence of diffusion encoding gradients. By repeating the MRI sequence with diffusion gradients applied along different directions, a diffusion tensor can be calculated for each image voxel. From this tensor, a set of three mutually perpendicular eigenvectors and eigenvalues can be calculated which describe the diffusion ellipsoid. In cardiac DTI (cDTI), the primary eigenvector (E1) corresponds to the local cardiomyocyte long-axis orientation, the secondary eigenvector (E2) corresponds to the local within-sheetlet cross-cardiomyocyte orientation, and the third eigenvector (E3) is perpendicular to E1 and E2, and therefore the sheetlet plane. From these data, it is possible to derive measures of myocardial tissue integrity, such as mean diffusivity (MD), fractional anisotropy (FA), E1A as an index of the mean intravoxel HA, and E2A as an index of the mean intravoxel sheetlet angle (SA)[18, 25].

The technical challenge for in vivo cDWI and cDTI is to detect incoherent diffusional motion on a micrometer scale in the setting of coherent bulk cardiac motion on a scale five orders of magnitude larger. Nonetheless, in vivo cDWI was successfully implemented for the first time by Edelman et al. [14]. Recent studies have demonstrated the potential of in vivo cDWI in detecting myocardial replacement fibrosis for chronic myocardial infarction [26, 27] and diffuse fibrosis in hypertrophic cardiomyopathy (HCM) [28]. Furthermore, cDTI methods have been reported to demonstrate the HA structure in the normal heart in vivo [15, 17] and in different pathological conditions [18, 19, 29-31], supported by studies validating ex vivo cDTI against histology [21-23]. cDTI data in ex vivo rodent hearts imaged separately in contracted and relaxed states supporting reorientation of laminar structures at different phases of the cardiac cycle has been reported [24, 25], as well as in vivo in healthy volunteers at systole [20].

Both intra-centre and inter-centre reproducibility studies of a quantitative technique for in vivo cardiac DTI have been preformed in healthy volunteers [32, 33] and in patients with HCM [34]. Using this technique, E2A changes from diastole to systole were presented which were hypothesised to represent dynamic rearrangement of sheetlets in healthy volunteers, as well as E2A changes consistent with systolic hyper-contraction and attenuated diastolic relaxation in patients with HCM [18]. No formal validation of in vivo cDTI against in situ cDTI and ex vivo cDTI together with paired histology data has been presented to date.

The aims of this study were to validate in vivo cDTI measures of cardiac microstructure against histology as well as to non­invasively characterise the microstructural dynamics underlying LV wall thickening in the loaded beating heart in vivo. This hypothesis was tested in a swine model, providing the first report of in vivo cDTI at 6-9 time points throughout the cardiac cycle, followed by in situ cDTI, ex vivo cDTI and co­registered histology in two contractile states. This design of these experiments enabled evaluation of myocardial microstructural dynamics in the presence and absence of bulk motion and strain.

Animal procedures were approved by the National Heart, Lung, and Blood Institute (NHLBI) Animal Care and Use Committee. In summary, in vivo cDTI was performed in Yorkshire pigs (N=16) at two mid-ventricular short axis (SAX) slices and at 6 to 9 time points across the cardiac cycle. Next, a mid-ventricular SAX slice was continuously imaged with cDTI in the intact animal in situ during the first hour after induction of cardiac arrest in both relaxed (N=6) and contracted (N=8) states by intravenous administration of potassium chloride (KCl) [24] and barium chloride (BaCl2) [24], respectively. The BaCl2 arrested hearts initially approximated a relaxed configuration but about 20-40 minutes after injection underwent one single final contraction over a time period of 5-10 minutes. This experiment effectively slowed cardiac contraction by three orders of magnitude (from ~300 ms in vivo in a beating heart to ~5 minutes) and enabled investigation of contractile motion by cDTI in the absence of cyclical strain effects. The hearts were then excised and imaged by ex vivo cDTI (N=16), after which tissue samples were acquired for paired histology (N=16).

The results demonstrated small changes in the cDTI measures of helix angle over the cardiac cycle. On the other hand, sheetlet angle cDTI measures changed substantially over the cardiac cycle. cDTI measures of sheetlet angle correlated significantly with wall thickness both in vivo and in situ. These changes with contraction were consistently observed under all experimental conditions and were in close agreement with quantitative histological results in both relaxed and contracted states.

This study shows that helical and laminar microstructures in the myocardium and their dynamic reorientations during cardiac contraction can be studied by in vivo cDTI non-invasively and non-destructively. Furthermore, it demonstrates in the loaded and beating heart in vivo that sheetlet reorientation is the predominant mechanism underlying myocardial LV wall thickening during systolic contraction. The results of this study support recent interpretation of in vivo cDTI data in patients with HCM, in whom maintenance of a systolic sheetlet orientation during diastole was suggested to account for increased LV wall thickness [18]. Further study of the microstructural dynamics of cardiac contraction and myocardial dysfunction with in vivo cDTI may produce new diagnostic and prognostic information in human cardiac disease.